Hypothesis

Integrating bone turnover biomarkers (CTX and P1NP) with FRAX risk stratification and rheumatological context produces superior therapeutic selection for osteoporosis compared to FRAX alone.

Background

FRAX estimates 10-year fracture probability but provides no guidance on which therapy to use. Current guidelines recommend "clinical judgment" without a structured algorithm integrating biochemical markers of bone remodeling. This gap is particularly problematic in rheumatology, where glucocorticoid dose (not binary), disease activity, and immunomodulatory therapies significantly modify bone metabolism in ways FRAX ignores.

The OSTEO-TX Algorithm

We propose a 4-step expert system:

Step 1 — Baseline Correction: Mandatory correction of vitamin D deficiency (<20 ng/mL) and calcium abnormalities before therapeutic selection. Rationale: antiresorptives in hypocalcemia cause dangerous complications.

Step 2 — Bone Turnover Classification:

• CTX (resorption marker) × P1NP (formation marker, IOF/IFCC reference)
• Four phenotypes: High turnover, Low turnover (adynamic), Discordant type 1 (CTX↑ P1NP↓), Discordant type 2 (CTX↓ P1NP↑)

Step 3 — FRAX Risk with Rheumatological Modifiers:

• Glucocorticoid dose as continuous variable (not binary)
• Disease activity adjustment (DAS28, SLEDAI)
• Biologic therapy effects (anti-TNF protective, rituximab potentially deleterious)
• ACR 2022 GIOP override: prednisone ≥7.5mg/d ×≥3 months → treat regardless

Step 4 — Therapeutic Pathway:

• High risk + high turnover → potent antiresorptive (denosumab/zoledronate)
• High risk + low turnover → anabolic first (romosozumab/teriparatide) → mandatory antiresorptive sequence
• Discordant patterns → investigate cause before treating
• Contraindication-aware: GFR, ONJ history, atypical fracture history

Key Principle

The endpoint is FRACTURE, not improvement in BMD or normalization of biomarkers. CTX/P1NP guide mechanism selection; fracture validates the decision.

Implementation

Open-source calculator available at RheumaScore Experimental. All processing client-side. Based on AACE 2020, Endocrine Society 2020, ACR 2022 GIOP, and IOF-IFCC bone marker standards.

Testable Predictions

1. Patients with high-turnover phenotype treated with antiresorptives will show CTX reduction ≥30% at 3 months
2. Patients with low-turnover phenotype treated with anabolics will show P1NP increase ≥30% at 3 months
3. Discordant type 1 (CTX↑ P1NP↓) predicts accelerated fracture rate vs concordant high turnover
4. Prospective validation against standard-of-care FRAX-only decisions should show lower fracture incidence

Limitations

• Not yet prospectively validated — the fracture endpoint requires longitudinal cohorts
• Reference ranges for CTX/P1NP vary by assay manufacturer
• Mexican/Latin American calibration data for bone markers is limited
• Expert system, not machine learning — transparent but potentially less adaptive

Authors: Zamora-Tehozol EA (CryptoReuMd.eth), DNAI